Configuration

PD

field	comments
name	The name of this pd
dataDir	The meta data storage dir
leaseSecsTTL	Lease time for PD leader
rpcAddr	Export grpc address for External components
embedEtcd:clientUrls	Etcd address for internal etcd client
embedEtcd:peerUrls	Etcd address for internal etcd peer
embedEtcd:initialCluster	PD cluster info
embedEtcd:initialClusterState	Iniial etcd cluster state
schedule:maxReplicas	The max replicas number for each cell
schedule:locationLabels	The location label for store, pd use this select best store for cell replicas
schedule:maxSnapshotCount	The max snapshot count for store, pd use this select best store for cell replicas
schedule:maxStoreDownTimeMs	The max time of store does not send heartbeat for PD. PD will not schedule cell replicas to this store
schedule:leaderScheduleLimit	The max count of leader cells at every store
schedule:cellScheduleLimit	The max count of cells at every store
schedule:replicaScheduleLimit	The max operator count of create cell replica at every store

Cell

command args	comments
clusterid	The Elasticell cluster id, 0 means not join the cluster
pd	PD addresses
addr	Internal address
addr-cli	KV client address
data	The data dir
zone	Zone label
rack	Rack label
buffer-cli-read	Buffer(bytes): bytes of KV client read
buffer-cli-write	Buffer(bytes): bytes of KV client write
batch-cli-resps	Batch: Max count of responses in a write operation
capacity-cell	Capacity(MB): cell
interval-heartbeat-store	Interval(sec): Store heartbeat
interval-heartbeat-cell	Interval(sec): Cell heartbeat
interval-split-check	Interval(sec): Split check
interval-compact	Interval(sec): Compact raft log
interval-report-metric	Interval(sec): Report cell metric
interval-raft-tick	Interval(ms): Raft tick
limit-peer-down	Limit(sec): Max peer downtime
limit-compact-count	Limit: Count of raft logs, if reach this limit, leader will compact [first,applied], otherwise [first, minimum replicated]
limit-compact-bytes	Limit(MB): Total bytes of raft logs, if reach this limit, leader will compact [first,applied], otherwise [first, minimum replicated]
limit-compact-lag	Limit: Max count of lag log, leader will compact [first, compact - lag], avoid send snapshot file to a little lag peer
limit-raft-msg-count	Limit: Max count of in-flight raft append messages
limit-raft-msg-bytes	Limit(MB): Max bytes per raft msg
limit-raft-entry-bytes	Limit(MB): Max bytes of raft log entry
threshold-compact	Threshold: Raft Log compact, count of [first, replicated]
threshold-split-check	Threshold(bytes): Start split check, bytes that the cell has bean stored
threshold-raft-election	Threshold: Raft election, after this ticks
threshold-raft-heartbeat	Threshold: Raft heartbeat, after this ticks
batch-size-proposal	Batch: Max commands in a proposal
batch-size-sent	Batch: Max size of send msgs
worker-count-sent	Worker count: sent internal messages
worker-count-apply	Worker count: apply raft log
enable-metrics-request	Enable: request metrics
enable-sync-raftlog	Enable: sync to disk while append the raft log
metric-job	prometheus job name
metric-address	prometheus proxy address
interval-metric-sync	Interval(sec): metric sync

Proxy

field	comments
addr	Address to serve redis
pdAddrs	GRPC address for PD cluster
maxRetries	Max times for retry send msg to backend servers
retryDuration	Duration interval for retry
supportCMDs	Support redis commands